On Wed, Apr 18, 2018 at 9:37 PM, Hendrik Leppkes <h.lepp...@gmail.com> wrote:
> ---
> libavformat/tls_schannel.c | 10 ++++++----
> 1 file changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/libavformat/tls_schannel.c b/libavformat/tls_schannel.c
> index 9a6e0c92e3..f41b007773 100644
> --- a/libavformat/tls_schannel.c
> +++ b/libavformat/tls_schannel.c
> @@ -413,11 +413,13 @@ static int tls_read(URLContext *h, uint8_t *buf, int
> len)
>
> ret = ffurl_read(s->tcp, c->enc_buf + c->enc_buf_offset,
> c->enc_buf_size - c->enc_buf_offset);
> - if (ret < 0) {
> + if (ret == AVERROR_EOF) {
> + c->connection_closed = 1;
> + ret = 0;
> + } else if (ret < 0) {
> av_log(h, AV_LOG_ERROR, "Unable to read from socket\n");
> return ret;
> - } else if (ret == 0)
> - c->connection_closed = 1;
> + }
>
> c->enc_buf_offset += ret;
> }
> @@ -515,7 +517,7 @@ cleanup:
> if (ret == 0 && !c->connection_closed)
> ret = AVERROR(EAGAIN);
>
> - return ret < 0 ? ret : 0;
> + return ret < 0 ? ret : AVERROR_EOF;
> }
>
> static int tls_write(URLContext *h, const uint8_t *buf, int len)
> --
> 2.17.0.windows.1
>
Ah yes, `retry_transfer_wrapper` which can return AVERROR_EOF :)
(which is what `ffurl_read` leads to). Funny how we have to convert
between zero and EOF upon entering and exiting due to the function's
initial design, but in general this gets a LGTM from me.
Jan
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
